Clustering Customers’ Behavior of an Online Store Offering e-learning Courses Using Machine Learning

Clustering customer behavior is very important issue for improving customer relation-ship management, including online store offering e-learning courses. The aim of the paper is to develop a method for clustering customers’ behavior of an online store offering e-learning courses using machine learning. The main contribution of the research is performing the clustering in the context of retention of customers’ behavior of an online store offering e-learning courses. The K-Means, self-organizing map (SOM) and Decision tree methods have been used in this research. The research results demonstrated that the primary factors influencing customer segment membership are the recency and monetary score columns, suggesting these should be examined in promotion profiling. Another key finding is that the maximum amount of money spent does not increase indefinitely with the number of purchases. It stays the same after 2 purchases. Conversely, the minimum amount spent rises with the number of purchases, indicating that customers are inclined to buy more courses at discounted prices and are less interested in paying full price after their initial purchase. Each of the conducted analyses suggests that the most important discriminating factor is the type of sales channel. Therefore, as the implication for practice, it seems reasonable to suggest that the price proposition algorithm should take it into account as one of the most important factors.